All AI News
    Discovery — Broad market AITuesday, September 1, 2026 14 min read
    AI

    EU Signs €387.8M LUMI-AI Supercomputer Contract Powered by AMD Instinct MI430X GPUs

    Europe's €387.8M LUMI-AI supercomputer bet on AMD signals a credible rival to Nvidia in sovereign AI infrastructure.

    Key takeaways
    • 01The EU has contracted a €387.8M expansion of the LUMI supercomputer, deploying AMD Instinct MI430X GPUs to create a dedicated AI research system.
    • 02The deal reinforces AMD's push into large-scale public-sector HPC at a moment when European governments are prioritizing homegrown AI capacity.
    • 03For AMD, a marquee European win at this scale challenges Nvidia's dominance in flagship AI infrastructure and could anchor future procurement decisions across member states.
    Koko brief

    Europe's €387.8M LUMI-AI supercomputer bet on AMD signals a credible rival to Nvidia in sovereign AI infrastructure.

    The EU has contracted a €387.8M expansion of the LUMI supercomputer, deploying AMD Instinct MI430X GPUs to create a dedicated AI research system. The deal reinforces AMD's push into large-scale public-sector HPC at a moment when European governments are prioritizing homegrown AI capacity. For AMD, a marquee European win at this scale challenges Nvidia's dominance in flagship AI infrastructure and could anchor future procurement decisions across member states.

    Watch: Whether AMD's MI430X performance benchmarks on LUMI-AI influence pending national AI supercomputer tenders in Germany, France, and the UK.

    Select any text and click on the icon to listen! gspeech.ioByGSpeech Since 1987: Covering the Fastest Computers in the World and the People Who Run Them TCI Media: Home of the Wire Publications hpcwire.comhpcwire.comhpcwire.com hpcwire.com SUBSCRIBE Shares [](facebook.com "Share this on Facebook")[](twitter.com "Tweet this !")[](linkedin.com "Add this to LinkedIn")[](reddit.com "Submit this to Reddit")[](news.ycombinator.com "Share this on HackerNews")[](mailto:?subject=HPCwire&body=Since%201987%20-%20Covering%20the%20Fastest%20Computers%20in%20the%20World%20and%20the%20People%20Who%20Run%20Them%20-%20https%3A%2F%2Fwww.hpcwire.com "Email this ")[](hpcwire.com "More share links") javascript:void(0) Off the Wire Press Releases September 1, 2026 NSF Establishes NAIRR Operations Center Led by SDSC and TACC MLCommons Releases New MLPerf Storage v3.0 Benchmark Results Brookhaven Lab to Lead $14M AI Project for Nation’s Electric Grid ALCF Service-Enabled Science Series Explores Experiment-Time Computing and the DIII-D Digital Twin Mixx Launches SxC Connector for High-Radix Scale-Up and Multi-Petabit Connectivity Amazon EC2 R9g and R9gd Instances Powered by AWS Graviton5 Now Generally Available Argonne AI Roadshow Connects Researchers with Local Communities QC Ware and IonQ Demonstrate High-Precision Hybrid Quantum Workflow for Drug Discovery August 31, 2026 TACC-Supported ASTRID Completes Largest Yet Cosmological Hydrodynamic Simulations Diraq to Deploy a Quantum Computer Inside an Equinix Data Center View All Off the Wire hpcwire.com Science & Technology Genesis Mission Takes on Astronomy’s Growing Computing Challenge by Ali Azhar The recently launched Genesis Mission includes a wide range of projects aimed at using advanced computing and AI to tackle some of the biggest challenges in science. One of those… hpcwire.com Community August 2026 HPC Career Notes by Alex Woodie It’s time once again for HPC Career Notes, our monthly feature that’s designed to keep you up-to-date on the latest career developments for individuals in the HPC community, including promotion,… * * hpcwire.com Datacenter Data Centers Are Under Attack. Is the Anger Justified? by Alex Woodie Anger over the construction of AI data centers has triggered political leaders to call for bans and moratoriums on the construction of new data centers in recent months. Is this… hpcwire.com Business Nvidia to Nab Hugging Face, the ‘GitHub for AI,’ for $12.9B: Report by Alex Woodie GPU giant Nvidia has agreed to buy Hugging Face, an AI model, data, collaboration and platform with more than 13 million registered users, for $12.9 billion, according to a report… hpcwire.com Life Sciences Nobel Laureate David Baker Takes Aim at the Virtual Cell with GenBio AI What if scientists could test a new drug or make changes to a cell without even touching one in the lab?... hpcwire.com semiconductor Inside Intel's 'Fan-out Fabric' in 'Diamond Rapids,’ Its Seventh-Gen Xeon Chip This week at Hot Chips, Intel shared more details of its upcoming seventh-generation Xeon chip, codenamed “Diamond Rapids,” which will utilize... hpcwire.com Nominations Open for Readers’ Choice Awards Nominations for HPCwire 2026 Readers’ Choice Awards are now open!Click to nominate your favorite HPC individuals, projects, or technologies for 2026. * * hpcwire.com Inside the Next Generation of Optical Interconnects The future of optical interconnects is very bright, thanks to the advantages they hold over copper networking in terms of bandwidth, reach, efficiency, and signal... hpcwire.com It’s Not an HPC System, But Cerebras’ New CS-4 Is an AI Monster Cerebras builds the biggest, most powerful chips in the industry with its Wafer Scale Engine 3, and so it’s not surprising to see that the... hpcwire.com What Hyperscalers Should Know About CXL The AI infrastructure conversation has shifted as AI inference becomes a larger part of workloads around the world. The cost of compute is no longer... hpcwire.com Scaling Up, Out, and Across: Interconnect Advice from a Meta Network Guru Meta is no stranger to large scale computing, with more than 3.5 billion daily users and one of the top AI labs in the world.... hpcwire.com What’s Hot Now: Custom AI Chips The AI boom has largely been powered by standardized hardware, including powerful GPUs, all-purpose CPUs, and as much HBM and DRAM as one can comfortably... hpcwire.com Deep Origin Claims Breakthrough in AI Drug Discovery Platform Deep Origin this month announced that its AI drug discovery framework delivered nearly a 31% hit rate for a cancer immunotherapy enzyme, nearly 100x higher... Gartner Finds Just 22% of Organizations Have Scaled AI Across Business Units DeepInfra and humans& Partner on 1,000-GPU NVIDIA B300 Cluster Data Centers Are Under Attack. Is the Anger Justified? Cost per Token vs. Cost per Task: What Really Determines AI ROI JUNIQ: Europe’s Quantum Buffet What the First Genesis Mission Projects Reveal About DOE’s AI Priorities Most Popular Articles hpcwire.com DeepSeek Open-Sources the Missing Layer Between AI Models and Agents AI models are getting better at a rapid pace. They are now able to reason, code and work through long… hpcwire.com Memory Shortages Are Rewriting the Rules of AI Inference The tech industry is fixated on one main metric, the raw number of GPUs accumulated by hyperscaler enterprises and governments.… hpcwire.com New GPU Method Compresses Scientific Simulation Data at 60 GB/s Scientists running large simulations can end up with terabytes of data that then has to be moved, stored and eventually… hpcwire.com FDEs are Tough to Find, But Essential to Develop OpenAI, Anthropic, Microsoft, AWS and others have recently launched initiatives to help their customers adopt AI and make sure it… More Most Popular hpcwire.com DeepSeek Open-Sources the Missing Layer Between AI Models and Agents AI models are getting better at a rapid pace. They are now able to reason, code and work through long… [](hp

    Don't miss tomorrow's

    The Daily Pulse in your inbox each morning — sourced and linked.

    How often
    Keep going — across the app